> For the complete documentation index, see [llms.txt](https://trust.impact.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://trust.impact.com/ai-trust-center/accountability-and-human-oversight.md).

# Accountability & Human Oversight

This page relates to our in-platform AI features. AI is a powerful tool, but people are ultimately accountable for how it's used. We build meaningful human oversight into every AI feature and take responsibility for our systems and their impact across the entire AI supply chain.

{% hint style="success" %}

### Humans stay in control

* **You decide.** Our AI recommends, surfaces insights, and drafts content, but meaningful decisions stay with people. There are clear pathways to escalate to a human for review.
* **Human-in-the-loop.** We keep humans in-, on-, and off-the-loop as appropriate, including people manually auditing AI interactions.
  {% endhint %}

## How we evaluate quality, reliability, and fairness

<table data-view="cards"><thead><tr><th></th><th></th></tr></thead><tbody><tr><td><strong>🤖 Automated testing &#x26; audits</strong></td><td>We run automated tests that compare AI answers against expected results to catch drift and quality regressions, and automated systems continuously review AI interactions to improve the product experience.</td></tr><tr><td><strong>👍 User feedback</strong></td><td>In-product feedback mechanisms (such as thumbs up/down) let you flag issues, which feed directly into our improvement process.</td></tr><tr><td><strong>🔍 Manual audits</strong></td><td>Our teams routinely review AI interactions and user feedback to monitor performance and catch problems automated checks might miss.</td></tr></tbody></table>

## Fairness & equity

We're dedicated to building AI that treats people fairly. We design for inclusion and accessibility, curate representative data, and test features for equitable performance across different user groups. Detailed fairness and bias audits are part of our testing process. See [Responsible by Design](/ai-trust-center/responsible-by-design-our-methodology.md).

{% hint style="info" %}
**How MCP differs.** In-platform AI is built around recommendations you review. MCP is built around authorized access: impact.com returns tool results to your AI client; it does not generate responses or decisions for you. You approve each client connection, and we enforce your existing permissions and audit every call.
{% endhint %}

## Regulatory adherence & end-to-end accountability

We hold our AI to global legal and regulatory standards, and we hold our third-party providers to the same governance principles we hold ourselves. Accountability doesn't stop at our platform boundary; it extends across every vendor in the AI supply chain.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://trust.impact.com/ai-trust-center/accountability-and-human-oversight.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.